Y603 Lectures
Online
Lecture #23
1. The within-subject assumption required by SPF
designs
Can understand the compound
symmetry or the circularity (or alas, the sphericity)
assumption.
When the above assumption is violated,
you know your three options.
Can conceptually present the pros and
cons of univariate versus multivariate approaches to the SPF data
analysis.
Can recognize an S or H matrix based on
the variance/covariance of repeated measures.
Know where to look in order to find
polynomial orthogonal coefficients needed in the multivariate
treatment of the SPF data.
2. Assignments:
(1) Review Section 12.4 in
Kirk.
(2) Review the article by McCall and
Appelbaum reproduced in your handout.
(3) Examine the sphericity
assumption for data presented in Questions 7 and 23 of Chapter 12
according to a SPF p.q design. Interpret the results
from SAS.
(4) Is there a need to examine
the sphericity assumption for Questions 8, 25, or 27 of Chapter
12?
(5) Preview Section 12.8 in
Kirk.
3. Announcement:
The second article critique is due next
Wednesday in class.
